Several Studies of the knee confirm that lower extremity weakness is associated with osteoarthritis (OA) of the knee, but it has not been clear whether this relationship extends to those with radiographic OA who do not have joint pain or whether the weakness precedes development of OA or is a result of the disease. Our ongoing study of 465 subjects over the age of 65, recruited from a random population sample of central Indiana, has identified quadriceps weakness (measured by isokinetic dynamometry) as a central feature of both radiographic and symptomatic OA in women, although men with OA demonstrate such weakness only when they also report pain. Studies of body composition in these subjects have shown that, despite being weaker, those with OA have significantly greater lean tissue mass (i.e., muscle) than those without OA. These data suggest that those with OA may have the potential to greatly increase their strength and thereby diminish their risk of developing more severe pain and disability or radiographic progression of OA. To test this possibility we now propose a randomized clinical trial of lower extremity strength training in four subgroups of subjects: l) radiographic OA with knee pain; 2) radiographic OA without knee pain; 3) knee pain without radiographic OA; and 4) normal elderly.